INDEX
    Explanations
    New Auto-Interp
    Negative Logits
    准确
    -0.08
    知识
    -0.08
    显示
    -0.08
    媒体
    -0.08
     encycl
    -0.08
     Datensch
    -0.08
     dividends
    -0.08
    按钮
    -0.08
     CDN
    -0.08
     accurate
    -0.08
    POSITIVE LOGITS
     anne
    0.11
     Acceptance
    0.09
     Cooling
    0.08
     aceitar
    0.08
     Ritual
    0.08
     Stove
    0.08
     cooled
    0.08
    %i
    0.08
     Anne
    0.08
     Magnum
    0.08
    Act Density 0.001%

    No Known Activations